Breaking

Anthropic Launches Claude Sonnet 5.5 With Faster Output and Lower Task Costs

September 28, 2026 at 14:37 EDT

Foundation Models
Software Dev & Coding
AI Agents

Anthropic released Claude Sonnet 5.5 on September 28, positioning the model as a faster, lower-cost workhorse for everyday coding, agent and knowledge-work tasks. The company says the new Sonnet is more than 30% faster than Claude Sonnet 5, costs up to 30% less per task in typical work, and is now available across the Claude Platform and major cloud partners.

September 28, 2026 - Anthropic

Claude Sonnet 5.5 shifts speed and cost

The new Sonnet model targets everyday coding, agent and knowledge-work tasks with faster output, lower task cost and tighter safeguards.

30%+

faster output than Sonnet 5

30%

lower task cost in typical work

55.5%

CursorBench 4.0 result reported by an early tester

$2/$10

input and output price per 1M tokens

Published coding comparison

CursorBench 4.0 puts Sonnet 5.5 close to Opus 5.5 while using a lower-priced tier.

55.5%

Sonnet 5.5

57.8%

Opus 5.5

Why teams like it

Early users reported faster codebase understanding, fewer tool calls, fewer output tokens and smoother iteration on scoped tasks.

Where caution remains

Anthropic still points complex open-ended work to Opus 5.5, and developers may need migration changes for thinking, tool choice and safeguards.

Deployment tradeoff

Routine coding>Lower effort>Lower latency and cost

Higher effort remains available when a task needs more checking, but the new Sonnet is positioned as the volume model for everyday agent work.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.